Google's shouting loudly about it's new almost-as-fast-as-you-think Instant search powers, and it is indeed clever. But a programmer worked out how to do this for Microsoft a year ago. Could Bing have beaten Google?
The clever geeky search guru we're talking about here is Long Zheng. A year ago, he wrote a front-end engine for Microsoft's own search engine, using existing APIs and coding tricks, that does almost exactly the same sort of "updating as you type" tricks that Instant does on Google. Check it out by clicking this link, and entering any old search phrase that comes to mind.
Impressive, isn't it? Notice how similar it feels to using Google's Instant powers. Zheng wrote it in his spare time too. The Web paid a modicum of attention to the invention, back in August 2009, and this week Microsoft's got all excited about it. Bing's director Stefan Weitz has even spoken to USA Today about it—noting that Google's Instant "interface itself isn't anything new. There was a site put up last year that does the same thing with Bing APIs." Then he talks some smack about how Bing is a "decision engine" (even though we all know it's no Wolfram Alpha!) and while speed and fewer clicks are all very good from a user convenience point of view, but the real game is "not about giving you much more links faster, it really is about getting you the information you need to make a decision faster in the format that makes the most sense." Stefan blithely ignores all the thinking about how much of a boost this will give to the online advert trade, but at least he does acknowledge that "speed is important."
The thing is, even if you think Instant is a gimmick, using it really is a transformational Web experience. It absolutely does make searching for something on the Intertubes rather surprising and possibly even fun. And it does reduce the effort of searching for something which, if you're in a online trade like mine, is rather welcome.
Weitz also acknowledges Google's genius in Instant was in being able to scale up the real-time search (using various predictive tricks) to provide the boosted service to millions of users without blowing up its server farms with the added burden.
But here's the problem: Though one can argue that Instant is "good" for Bing in some ways, Zheng provided the same idea to Bing and MS last year, effectively for free. And the company's exec team surely noticed. Then they ignored it. They've got thousands of smart programmers, huge server farms, and experts in search on staff. They too could've come up with clever ways to upscale the system for their millions of users, and totally beaten Google to the punch. But ... they didn't.